Detonator 51, 5900, and image quality

OK, much has been made recently about how NVidia doesn't run fast enough for
HL 2, that it won't look as good in DX 9, etc. I got a hold of the free
release Aquamark 3 and leaked 51.57 detonator drivers. I thought I'd
perform a test of my own.

First, the test systems:

Radeon 9700 Pro
Athlon XP 2000
NForce II motherboard
512 MB DDR RAM
19 inch CRT (no-name)
Catalyst 3.6 drivers (quality, 4xAA, maximum anisotropic in drivers)

GeForce FX 5900 (non-pro)
Athlon XP 2400
KT400 motherboard
512MB DDR Ram
17 inch NEC 1712 LCD.
Beta Detonator 51.75 (quality, 4xAA, maximum anisotropic in drivers)

Both are using Windows XP operating systems. The CRT is calibrated for a
fairly even greyscale gamut. The LCD is calibrated for optimal game image
quality with 3Deep. I don't own the Radoen 9700 Pro system, but I was able
to borrow it for the test.

Results for default test (very high quality, 1024X768):
system average FPS
Radeon 9700 Pro 24.6
GeForce FX 5900 25.8

So both cards are roughly equally fast. It's clearly a very demanding
benchmark.

Image quality

Both look very good. I really couldn't tell much difference between
either one. I saw no evidence of color banding with the GeForce FX 5900.
Mind you, I wasn't taking screenshots and dissecting them. If I had to make
a snap judgement, I'd say the Radeon 9700 Pro's image was a little more
vibrant, but maybe that's just because it was on a CRT.

I should also point out the LCD was running in a non-native resolution and
aspect ratio, but I still thought both screens looked roughly the same.

So this leaves me with an early impression that everybody badmouthing
NVidia is jumping the gun. Rumors that CompUSA and several OEM dealers
ditching NVidia, seems premature.

Now, a few caveats. I do like NVidia cards. I like ATI cards too, but
generally I like NVidia's drivers better. I've owned both in the past, and

I'll freely admit my bias is slightly towards NVidia, simply because they
RUN EVERY GAME I OWN with very little quality loss, and ATI, sadly cannot
live up to that claim.

Feedback anybody?

"magnulus" wrote in message
.. .
OK, much has been made recently about how NVidia doesn't run fast enough
for
HL 2, that it won't look as good in DX 9, etc. I got a hold of the free
release Aquamark 3 and leaked 51.57 detonator drivers. I thought I'd
perform a test of my own.

First, the test systems:

Radeon 9700 Pro
Athlon XP 2000
NForce II motherboard
512 MB DDR RAM
19 inch CRT (no-name)
Catalyst 3.6 drivers (quality, 4xAA, maximum anisotropic in drivers)

GeForce FX 5900 (non-pro)
Athlon XP 2400
KT400 motherboard
512MB DDR Ram
17 inch NEC 1712 LCD.
Beta Detonator 51.75 (quality, 4xAA, maximum anisotropic in drivers)

Both are using Windows XP operating systems. The CRT is calibrated for
a
fairly even greyscale gamut. The LCD is calibrated for optimal game image
quality with 3Deep. I don't own the Radoen 9700 Pro system, but I was able
to borrow it for the test.

Results for default test (very high quality, 1024X768):
system average FPS
Radeon 9700 Pro 24.6
GeForce FX 5900 25.8

So both cards are roughly equally fast. It's clearly a very demanding
benchmark.

Image quality

Both look very good. I really couldn't tell much difference between
either one. I saw no evidence of color banding with the GeForce FX 5900.
Mind you, I wasn't taking screenshots and dissecting them. If I had to
make
a snap judgement, I'd say the Radeon 9700 Pro's image was a little more
vibrant, but maybe that's just because it was on a CRT.

I should also point out the LCD was running in a non-native resolution
and
aspect ratio, but I still thought both screens looked roughly the same.

So this leaves me with an early impression that everybody badmouthing
NVidia is jumping the gun. Rumors that CompUSA and several OEM dealers
ditching NVidia, seems premature.

Now, a few caveats. I do like NVidia cards. I like ATI cards too, but
generally I like NVidia's drivers better. I've owned both in the past,
and

I'll freely admit my bias is slightly towards NVidia, simply because they
RUN EVERY GAME I OWN with very little quality loss, and ATI, sadly cannot
live up to that claim.

Feedback anybody?

You used a 9700Pro...i think that a 9800Pro should have been used as the
test card for comparison as most people will want to know how these two
cards perform against one another.

Anyone considering a purchase, me included, will be buying either one of
these cards.

"Richard Dower" wrote in message
...
You used a 9700Pro...i think that a 9800Pro should have been used as the
test card for comparison as most people will want to know how these two
cards perform against one another.

Good idea. Would you like to donate 300 dollars so I could procure one?

Seriously, a Radeon 9700 Pro is a DX 9 card and not very different from a
Radeon 9800 Pro in terms of image quality, which was the main focus of my
test.

And BTW, my conclusion is Gabe Newel, until he shows me the screenshots,
is bought and paid by ATI. Sad, because I really like Half Life.

"magnulus" wrote in message
.. .

"Richard Dower" wrote in message
...
You used a 9700Pro...i think that a 9800Pro should have been used as the
test card for comparison as most people will want to know how these two
cards perform against one another.

Good idea. Would you like to donate 300 dollars so I could procure one?

Seriously, a Radeon 9700 Pro is a DX 9 card and not very different from
a
Radeon 9800 Pro in terms of image quality, which was the main focus of my
test.

Well you said the fps were almost identical, so a higher clocked 9800Pro
should get even more fps over the FX5900np. I'd still like to see an FX
Ultra against the 9800Pro though.

Some more observations:

ATI's Animusic Pipe-Dream R 9700 Pro demo actually looks better running on
the GeForce FX 5900 with the new Det. 51 drivers than it did with the Det
45's. It corrects some missing polygons that weren't there before (the rail
some xylophone slides along). Still, it runs MUCH slower than on the Radeon
9700 Pro. I assume the demo is optimized for ATI cards. ATI's cel-shaded
demo with the wine glasses and fruit runs reasonably good on the GeForce FX-
you just have to tweak it in the menu slightly otherwise the screen is
black.

Also, the Dusk Ultra, if it represents anything like a finished game,
doesn't run very well with the Det 45's. The performance is only a little
better than a slideshow- I can't imagine making a game with NVidia cards and
expecting it to look that quality and getting it to run well. I can't even
get it to run properly with the Det. 51 drivers. Dawn Ultra is also slow,
but slightly better. Antialiasing doesn't seem to affect the framerate
much, if at all. "Last Chance Gas" runs very slowly too, and I may have
seen some color banding in the sky (which is procedurally generated)- but ag
ain I'm running on an LCD.

so you think that the cpu is not holding back the 9700 ?
the 1.2 fps diferance is mostly due to cpu.
next time you test try swaping the cards in the same sys and do fresh
install of xp on each set up then defrag and all that then test .
hell if i wanted to make the 9700 look better than the 5900 all i have to do
is swap the cpu's and run the 9700 with the 2400xp and the 5900 on the
slower 2000xp.
im not trying to **** you off but if you want to test and be fair then do it
the right way.
and if you dont think that the cpu maters here is mine with the 9700pro and
a 2500+ cpu 34.62 fps
and if you dont belive me here is link
http://arc.aquamark3.com/arc/arc_vie...runID=21383831
hm the 2500+ is not a lot faster than the 2400+ that you used for the test
yet im getting 9 fps more with just 100mhz more cpu. and you tested with a
400 mhz diferance and got what 1.2 fps more?. are you getting my point yet?
well the 100mhz and 400mhz thing is not realy true due to amd nameing game
but you shuld get my point.
last years 9700 pro will beat a brand new this year fx5900 non pro in a fair
test. lets not even get in to the 9800. i take it the 5900 is yours so your
defending it lol. well i also have a fx but a cheap one the low end 5200 and
i like it "for what i use it for" but i dont defend it. hm i think i shuld
bench it against the 9700 but only if i crippel it and run the 9700 with a
slower cpu like one of my old amd 700 cpu and then compare it to a 3200+ cpu
with the 5200 with drivers that make it look good.
if you want to compare the 9700 then bench it against what it was made to
run against the 5800 leaf blower. oh my god im defending my 9700 pro lol.
im glad you tryed testing them against each other the best you could but
next time do a even test
--
If life is a bitch then live it like she is drunk and naked.

-------------------------------------------------------------------------
FIGHT BACK AGAINST SPAM!
Download Spam Inspector, the Award Winning Anti-Spam Filter
http://mail.giantcompany.com


"magnulus" wrote in message
.. .
OK, much has been made recently about how NVidia doesn't run fast enough
for
HL 2, that it won't look as good in DX 9, etc. I got a hold of the free
release Aquamark 3 and leaked 51.57 detonator drivers. I thought I'd
perform a test of my own.

First, the test systems:

Radeon 9700 Pro
Athlon XP 2000
NForce II motherboard
512 MB DDR RAM
19 inch CRT (no-name)
Catalyst 3.6 drivers (quality, 4xAA, maximum anisotropic in drivers)

GeForce FX 5900 (non-pro)
Athlon XP 2400
KT400 motherboard
512MB DDR Ram
17 inch NEC 1712 LCD.
Beta Detonator 51.75 (quality, 4xAA, maximum anisotropic in drivers)

Both are using Windows XP operating systems. The CRT is calibrated for
a
fairly even greyscale gamut. The LCD is calibrated for optimal game image
quality with 3Deep. I don't own the Radoen 9700 Pro system, but I was able
to borrow it for the test.

Results for default test (very high quality, 1024X768):
system average FPS
Radeon 9700 Pro 24.6
GeForce FX 5900 25.8

So both cards are roughly equally fast. It's clearly a very demanding
benchmark.

Image quality

Both look very good. I really couldn't tell much difference between
either one. I saw no evidence of color banding with the GeForce FX 5900.
Mind you, I wasn't taking screenshots and dissecting them. If I had to
make
a snap judgement, I'd say the Radeon 9700 Pro's image was a little more
vibrant, but maybe that's just because it was on a CRT.

I should also point out the LCD was running in a non-native resolution
and
aspect ratio, but I still thought both screens looked roughly the same.

So this leaves me with an early impression that everybody badmouthing
NVidia is jumping the gun. Rumors that CompUSA and several OEM dealers
ditching NVidia, seems premature.

Now, a few caveats. I do like NVidia cards. I like ATI cards too, but
generally I like NVidia's drivers better. I've owned both in the past,
and

I'll freely admit my bias is slightly towards NVidia, simply because they
RUN EVERY GAME I OWN with very little quality loss, and ATI, sadly cannot
live up to that claim.

Feedback anybody?

I'll freely admit my bias is slightly towards NVidia, simply because they
RUN EVERY GAME I OWN with very little quality loss, and ATI, sadly cannot
live up to that claim.

I just wish the ATI systems I own would simply run all of the games I have,
let alone running them well. As it is - half the games I own from the past
year need some kind of tweak or third-party driver to act correctly.

Can you name those games that don't work with a Radeon 9x00?


"TheHasselhoff" wrote in message
om...
I'll freely admit my bias is slightly towards NVidia, simply because
they
RUN EVERY GAME I OWN with very little quality loss, and ATI, sadly
cannot
live up to that claim.

I just wish the ATI systems I own would simply run all of the games I
have,
let alone running them well. As it is - half the games I own from the
past
year need some kind of tweak or third-party driver to act correctly.

magnulus wrote:

I'll freely admit my bias is slightly towards NVidia, simply because
they RUN EVERY GAME I OWN with very little quality loss, and ATI,
sadly cannot live up to that claim.

Feedback anybody?

Are you aware that 3dchipset pulled the 51.75 drivers--a mirror did, anyway.
Check it out
http://www.m3dzone.com/modules.php?n...ticle&sid=2649
--
chainbreaker

If you need to email, then chainbreaker (naturally) at comcast dot
net--that's "net" not "com"--should do it.

"Tim Miser" wrote in message
t...
Can you name those games that don't work with a Radeon 9x00?



Blade of Darkness, Rally Masters. There are probably others, just haven't
tested out every single game I own. In addition, the following games had
problems for a long time with ATI hardware (but finally the problem was
resolved):

Splinter Cell- you could see lights through walls, etc.
Metal Gear Solid II: Substance- totally fvcked
Deus Ex- detailed textures caused wierd banding

Only 1 game (Blade of Darkness) has any problem at all on NVidia hardware,
and the bug doesn't make the game unplayable.

It's not a big deal, and if you don't play those games, well, it doesn't
matter much. I just like NVidia's drivers better because when I get a game,
I know it'll at least run and won't be the graphic card's fault. Now,
maybe that'll change for DX 9 games, but honestly I didn't expect DX 9 games
to be out so soon.